Skip to content

Added cache_block_outputs option to enable GPTQ for non-regular models #27032

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 5 commits into from
Nov 1, 2023
Merged

Added cache_block_outputs option to enable GPTQ for non-regular models #27032

Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
4abc7f1
Added cache_block_outputs option to enable GPTQ for non-regular models
AlexKoff88 Oct 24, 2023
62eedeb
Update src/transformers/utils/quantization_config.py
AlexKoff88 Oct 31, 2023
8320cb7
Update src/transformers/utils/quantization_config.py
AlexKoff88 Oct 31, 2023
f9f72a9
Fixed style
AlexKoff88 Nov 1, 2023
027e184
Update src/transformers/utils/quantization_config.py
AlexKoff88 Nov 1, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 4 additions & 0 deletions src/transformers/utils/quantization_config.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -349,6 +349,8 @@ class GPTQConfig(QuantizationConfigMixin):
max_input_length (`int`, *optional*):
The maximum input length. This is needed to initialize a buffer that depends on the maximum expected input
length. It is specific to the exllama backend with act-order.
cache_block_outputs (`bool`, *optional*, defaults to `True`):
Whether to cache block outputs to reuse as inputs for the succeeding block.
"""

def __init__(
Expand All @@ -369,6 +371,7 @@ def __init__(
pad_token_id: Optional[int] = None,
disable_exllama: bool = False,
max_input_length: Optional[int] = None,
cache_block_outputs: bool = True,
**kwargs,
):
self.quant_method = QuantizationMethod.GPTQ
Expand All @@ -388,6 +391,7 @@ def __init__(
self.pad_token_id = pad_token_id
self.disable_exllama = disable_exllama
self.max_input_length = max_input_length
self.cache_block_outputs = cache_block_outputs
self.post_init()

def get_loading_attributes(self):
Expand Down